The Joint Commission requires tracking and reporting several clinical performance criteria, aptly called core measures. Focused on particular clinical entities, those relevant to emergency medicine include acute myocardial infarction, congestive heart failure, pneumonia, and asthma. Emergency physicians are most often responsible for three components of pneumonia treatment: oxygenation assessment, blood cultures, and antibiotics within four hours of hospital arrival, but nationwide performance of these measures is at only 79 percent.
Most departments that achieve high levels of performance on these measures do so with carefully crafted processes to identify patients with pneumonia quickly and use standardized order sets that administer antibiotics. The model starts with triggers at intake.
One study that may have escaped most EPs' attention was published in the Archives of Internal Medicine. (1997;157 :1453.) In this study, the authors stratified the prevalence of certain community-acquired pneumonia (CAP) clinical features by age group. The results showed an increased frequency of pleuritic chest pain, with almost 60 percent of patients under 45 complaining of this. Conversely, young patients were infrequently tachypneic compared with older patients. The point is that emergency physicians can use these clinical features to initiate an intake process to identify and expedite patients who may have pneumonia.
Some emergency departments have implemented successful strategies to improve on CAP core measures, including:
▪ Early identification of possible CAP: Based on triggers at intake, some departments put a colored order set for visual cueing on the chart. Others flag the chart to expedite x-rays, or have a system to move these patients to the front of the x-ray queue.
▪ Expedited x-ray reads: Just as the ED is alerted to the possibility of CAP, radiology needs to be alerted. Some institutions that have real-time radiology have set CAP as a “critical alert” with prompt notification of the ED. Very high-tech departments with real-time reads and voice recognition transcription have set cues on the tracking system to pop up when the transcription service types key words such as “infiltrate,” “haziness,” “pneumonia,” and “patchy.” This launches an immediate visual cue on the tracking system that pneumonia has been read by radiology.
One department in the ED Collaborative at the Institute for Healthcare Improvement (IHI) had experienced x-ray technicians working at night, and they empowered them to bring any x-ray with a hint of pneumonia to the emergency physician for an immediate wet read. This strategy greatly improved their time-to-diagnosis for pneumonia.
▪ Cueing for blood cultures: This is likely the most controversial element of the CAP core measures, and it is still being monitored by the Joint Commission. It probably improves care to have blood culture results on critically ill patients with pneumonia, and some interesting processes have evolved to improve compliance with blood culture draws prior to antibiotic administration. One IHI team places a colored armband on any patient with possible pneumonia in triage. If blood cultures are drawn, the band is torn off. If a nurse is about to administer antibiotics, she checks for the armband. If it is in place, she asks for cultures to be drawn before hanging the antibiotics. If there is no band in place, the nurse double-checks with the lab that cultures were sent before administering antibiotics. This low-tech strategy also yielded improvements at that institution.
▪ Antibiotic orders: Delays also might occur in the overall CAP process at antibiotic administration. Many departments have found success by crafting standardized order sets when pneumonia is recognized. When a real-time radiology read of pneumonia is made at Vanderbilt University, the physician cannot complete any actions on the computer for that patient until he approves or disapproves the pneumonia orders. This kind of IT support for ED processes will be part of our future, and assist us in these time-dependent tasks.
One success story that deserves telling is from Northeastern Vermont Regional Hospital in St. Johnsbury. This institution boasts 100 percent compliance with CAP core measures for nine of the past 10 months. In 2002, the hospital was reporting compliance rates below CAP core measures: 87 percent for documentation of oxygen saturation, 75 percent for blood cultures drawn before antibiotic administration, and 69 percent for antibiotic administration within four hours of hospital arrival.
The improvement team began with education and training for the ED staff, and then met with a committee representing the admitting physicians. They posted printed material to act as a reminder to staff, and the ED took ownership of the process and rapidly improved their compliance rates: 100 percent for documentation of oxygen saturation, 100 percent for blood cultures drawn before antibiotic administration, and 83 percent for antibiotic administration within four hours of hospital arrival.
Improvement continued, and the gains were held, despite one blip with compliance for antibiotic administration. That coincided with the launch of a new hospitalist program, which had caused confusion about who was writing the antibiotic orders. Once this problem was identified, the team worked with the ED and the hospitalists to craft a miniature order set for antibiotic administration that dovetailed with the existing admission orders for pneumonia, resulting in a return to high compliance.
The improvement team credited the feedback loop that provided compliance and performance data to staff members with having a huge impact on the process. Health care workers generally and physicians particularly are competitive and performance-driven. It takes very little to motivate them to reach performance goals. Northeastern Vermont also recommended providing positive and negative individualized feedback, being careful to use wording such as “opportunity for improvement” instead of “deficiency” or “failure” when a measure is not met.